Sherin Babu - Corresponding author: Department of Computer Science, Assumption College Autonomous, Changanacherry, Kottayam, Kerala, India
Binu Thomas - School of Computer Sciences, Mahatma Gandhi University, Kottayam, Kerala, India

Background: One of the most critical contributors to air pollution is particulate matter (PM۲.۵) that its acute or chronic exposure causes serious health effects to human. Accurate forecasting of PM۲.۵ concentration is essential for air pollution control and prevention of health complications. A survey of the available scientific literature on random forest model for PM۲.۵ prediction is presented here. Methods: The scientific literature is extracted from Science Direct database based on a set of specified search criteria. The input features, data length, and evaluation parameters used in PM۲.۵ prediction were analyzed in this study. Results: The study shows that majority of the publications are aimed at the daily prediction of outdoor PM۲.۵. Most publications base their PM۲.۵ prediction on features aerosol optical depth (AOD) and boundary layer height (BLH). PM۱۰ and NO۲ are the main air pollutants employed in the PM۲.۵ estimation. Majority studies utilized input data lengths covering more than one year, and the effectiveness of prediction models are unaffected by the length of investigation. The coefficient of determination, R۲, is the primary evaluation parameter used in all publications. The majority of research study indicated R۲ values greater than ۰.۸۵, demonstrating the reasonable dependability and efficiency of random forest regression-based PM۲.۵ prediction models. Conclusion: The study demonstrates that the publications use a variety of meteorological and geological features for PM۲.۵ estimation, depending on the context of the research as well as data accessibility. The findings demonstrate that it is hard to pinpoint the optimal model in any particular way.

Air pollution, Air pollutants, Aerosols, Particulate matter, Machine learning